Laboratory for Computation in Finance and Commerce
|
More Information: Welcome News Projects Publications |
 |
Projects
| COMMODITY TRADING | This project studies computation problems related to inventory management, and commodity trading. The focus is on efficient algorithms for evaluating and ranking various decision options, and for estimating portfolio valuation and risk.
(Project Lead: Eli Upfal) |
| TRANSPORTATION NETWORKS | Unreliable transportation networks pose a serious challenge for the reliable supply of resources. This projects investigates robust strategies for trading commodities such as electricity and natural gas which are provided through transportation networks.
(Project Lead: Eli Upfal and Thomas Hofmann) |
| MULTI AGENT LEARNING |
This project is concerned with multi-agent learning among Internet agents engaged in game-theoretic (i.e., strategic) activities, such as bidding in on-line auctions and dynamic pricing among shopbots and pricebots
(Project Lead: Amy Greenwald) |
| RECOMMENDER SYSTEMS | Recommender systems are used in electronic commerce to
help people to find products and information they are looking for. Our
research has been focused on statistical methods for collaborative or
social filtering that make recommendations based on known user profiles.
(Project Lead: Thomas Hofmann) |
| CONFIGURATION SYSTEMS | Configuration systems are used in electronic commerce to help design a complex product (e.g., a personal computer system) from components (e.g., RAM, network cards) given constraints on its functionality (e.g., 3-D graphics, high-speed communication) and, possibly, multiple decision criteria. Our research focuses on the design of algorithmic techniques and languages to support the design of configuration systems. (Project Lead: Pascal Van Hentenryck) |
Enabling Technologies
| PLANING UNDER UNCERTAINTY |
The focus of this research project is on planning under uncertainty
using Markov decision processes. The main application areas is the
design of automated planning and scheduling systems for stochastic
domains. The theoretical emphasis is on algorithms for solving Markov
decision processes with very large state and action spaces.
(Project Lead: Tom Dean) |
| STOCHASTIC OPTIMIZATION | In combinatorial optimization problems such as optimal
scheduling and resource allocation, the variables that define the
problem instance are typically not known to arbitrary precision in
advance due to measurement noise and system failures. This project
aims at developing optimization techniques and languages to compute
robust and alterable solutions to stochastic optimization problems.
(Project Lead: Pascal Van Hentenryck) |
 |
| Page Owner: Thomas Hofmann |
| Page Owner: Thomas Hofmann | Last Modified: Mon Apr 21 17:39:11 2003 |